| > a cohort is simply a defined group of people followed over time. Yes, and the group is defined very clearly here, as is usually the case. I appreciate that you've conceded that it does not include everyone, as we see in your next quote: > It used the hospitals EHR to identify the numerator (the heart inflammation cases) out of the denominator (the entire regional population). So it compared something studied in the cohort (the numerator) to something not studied and outside the cohort (the denominator). So now let's establish whether or not the denominator can be used without calling ourselves science deniers. And as we go, let's consider that a slight adjustment to that denominator has a multiplicative impact on the result! So we really want to get that dialed otherwise our interpretation could lie very, very far from the truth. Your next quote leads us right to the most convincing point about the denominator. > You are confusing PCR testing with serology. I said all methods measuring prevalence in 2020 were wrong. That doesn't imply confusion with PCR. Serology also was wrong. Just to find something agreeable to cite, here is the CDC stating very clearly that we shouldn't use serology for this kind of decision making. [0] "Negative results do not rule out SARS-CoV-2 infection and should not be used as the sole basis for treatment or patient management decisions, including infection control decisions." So, assuming you trust the CDC publishes good science, then you know damned well you can't trust that denominator you keep throwing around as fact. At the very least, you're in opposition to CDC guidance if you want to keep using that denominator. You're not a science denier are you? > This is... literally what I've been doing the entire time. Well, you may want to read back. The claim that a cohort includes everyone is not remotely in the realm of science. Neither of us can conceive of an experiment that would obtain the denominator to be used in the math above. Not as it is, and definitely not if you defined it as "everyone else." And that's been my point for the last few comments. It's an illogical, fantasy based approach to interpreting the data. You did not test everyone, and you never will for any research ever in the past or future. It's not even fathomably possible. The denominator, implicitly, makes broad assumptions about cultural, environmental, and economic factors, and that's just the beginning. There are countless other factors that we haven't even thought of yet. So that denominator, can be seen as a wild guess at best, and at worst it's a blatant falsehood if the study was published after whatever research the CDC is relying on to tell us not to use it. And therefore, it must be considered with all of the ambiguity that it implicates, as a wild guess or a blatant falsehood. I don't like to take things that are wild guesses and stuff them into my math. There's a special word for that, pseudoscience. I have more respect for science than to do that. So, all things considered, I'll take your word for it that sticking to the science is literally what you've been doing. And you can literally do it better by not using wild guesses to try to make the math say what you want. If you will use numbers that are based on measured (or even measurable!) data, I'll be more inclined to lend belief to what insight you have that's worth exploring. And I hope you will. -- Summary: If you have an argument that makes sense, I'll gladly follow it. But you're not there yet. Your argument is entirely based on a denominator that doesn't remotely meet what modern philosophers of science would refer to as worth believing in. It's a wild guess that the CDC says is inaccurate for weighing this kind of decision. [0] https://www.cdc.gov/covid/hcp/clinical-care/overview-testing... |